Skip to content
Back to search

claude-for-crypto-research

This is a simple implementation for crypto research agent inspired by Claude Skills repo

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/juntoku9--claude-for-crypto-research/badge.svg)](/listings/juntoku9--claude-for-crypto-research)

HTML

<a href="/listings/juntoku9--claude-for-crypto-research"><img src="/api/listings/juntoku9--claude-for-crypto-research/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

# Install dependencies
uv sync

# Set your API keys
export ANTHROPIC_API_KEY="your-anthropic-api-key"
export TWITTER_API_KEY="your-twitter-api-key"

# Or copy .env.example to .env and fill in your keys
cp .env.example .env

# Run the agent (choose mode)
uv run python research_agent/agent.py --mode fast   # Fast mode (2-3 min)
uv run python research_agent/agent.py --mode deep   # Deep mode (8-12 min, default)
uv run python research_agent/agent.py               # Default is deep mode

Then ask: "Research Arbitrum" or "Research OpenMind token"

API Keys Required

- **Anthropic API Key**: Get from [console.anthropic.com](https://console.anthropic.com/settings/keys) - **Twitter API Key**: Get from [twitterapi.io](https://twitterapi.io) for Twitter/X sentiment analysis

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 152 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # Crypto Research Agent
  2. A multi-agent crypto research system that coordinates specialized subagents to research crypto projects, tokens, DeFi protocols, and market trends. Generates comprehensive Markdown reports with airdrop participation guides, market analysis, and Twitter sentiment.
  3. ## Quick Start
  4. ```bash
  5. # Install dependencies
  6. uv sync
  7. # Set your API keys
  8. export ANTHROPIC_API_KEY="your-anthropic-api-key"
  9. export TWITTER_API_KEY="your-twitter-api-key"
  10. # Or copy .env.example to .env and fill in your keys
  11. cp .env.example .env
  12. # Run the agent (choose mode)
  13. uv run python research_agent/agent.py --mode fast # Fast mode (2-3 min)
  14. uv run python research_agent/agent.py --mode deep # Deep mode (8-12 min, default)
  15. uv run python research_agent/agent.py # Default is deep mode
  16. ```
  17. Then ask: "Research Arbitrum" or "Research OpenMind token"
  18. ### API Keys Required
  19. - **Anthropic API Key**: Get from [console.anthropic.com](https://console.anthropic.com/settings/keys)
  20. - **Twitter API Key**: Get from [twitterapi.io](https://twitterapi.io) for Twitter/X sentiment analysis
  21. ## Research Modes
  22. ### Fast Mode (2-3 minutes)
  23. ```bash
  24. uv run python research_agent/agent.py --mode fast
  25. ```
  26. - **Speed**: 2-3 minutes per research
  27. - **Researchers**: 1-2 agents spawned
  28. - **Searches**: 3-4 essential searches per researcher
  29. - **Twitter**: 10 high-quality tweets (50+ likes, verified)
  30. - **Coverage**: Essentials (overview, price, sentiment, airdrops)
  31. - **Use case**: Quick overview, price check, airdrop opportunities
  32. ### Deep Mode (8-12 minutes) - DEFAULT
  33. ```bash
  34. uv run python research_agent/agent.py --mode deep
  35. uv run python research_agent/agent.py # Same as deep
  36. ```
  37. - **Speed**: 8-12 minutes per research
  38. - **Researchers**: 3-4 agents spawned in parallel
  39. - **Searches**: 6-8 comprehensive searches per researcher
  40. - **Twitter**: 15-20 tweets with broader coverage
  41. - **Coverage**: Comprehensive (tech, tokenomics, TVL, competitive analysis)
  42. - **Use case**: Investment decisions, thorough due diligence
  43. ## How It Works
  44. 1. **Lead Agent** breaks your request into 2-4 subtopics
  45. 2. Spawns **Researcher** subagents in parallel to search the web and Twitter
  46. 3. Each Researcher saves findings to `files/research_notes/`
  47. 4. Spawns **Report Writer** to create final Markdown report in `files/reports/`
  48. ## Agents
  49. | Agent | Tools | Purpose |
  50. |-------|-------|---------|
  51. | **Lead Agent** | `Task` | Coordinates research, delegates to subagents |
  52. | **Researcher** | `WebSearch`, `Write`, `Skill`, `Bash` | Gathers info from web and Twitter, uses site-specific searches |
  53. | **Report Writer** | `Write`, `Glob`, `Read` | Creates Markdown reports with participation guides |
  54. ## Slash Commands
  55. | Command | Description |
  56. |---------|-------------|
  57. | `/research <topic>` | Start focused research on any topic |
  58. | `/competitive-analysis <company>` | Analyze companies or products |
  59. | `/market-trends <industry>` | Research industry trends |
  60. | `/fact-check <claim>` | Verify claims and statements |
  61. | `/summarize` | Summarize all current research findings |
  62. ## Example Queries
  63. **Fast Mode** (quick checks):
  64. - "Research OpenMind token"
  65. - "What's the sentiment on Arbitrum?"
  66. - "Check if there's an airdrop for Gradient"
  67. **Deep Mode** (thorough research):
  68. - "Research Arbitrum L2 ecosystem"
  69. - "Compare Aave vs Compound DeFi protocols"
  70. - "Research the AI x Crypto narrative"
  71. **Slash Commands** (any mode):
  72. - `/research Solana`
  73. - `/competitive-analysis Uniswap`
  74. - `/market-trends DeFi`
  75. ## Output Structure
  76. ```
  77. files/
  78. ├── research_notes/ # Markdown files from researchers
  79. └── reports/ # Final Markdown reports with participation guides
  80. logs/
  81. └── session_YYYYMMDD_HHMMSS/
  82. ├── transcript.txt # Human-readable conversation
  83. └── tool_calls.jsonl # Structured tool usage log
  84. ```
  85. ## Subagent Tracking with Hooks
  86. The system tracks all tool calls using SDK hooks.
  87. ### What Gets Tracked
  88. - **Who**: Which agent (RESEARCHER-1, DATA-ANALYST-1, etc.)
  89. - **What**: Tool name (WebSearch, Write, Bash, etc.)
  90. - **When**: Timestamp
  91. - **Input/Output**: Parameters and results
  92. ### How It Works
  93. Hooks intercept every tool call before and after execution:
  94. ```python
  95. hooks = Hooks(
  96. pre_tool_use=[tracker.pre_tool_use_hook],
  97. post_tool_use=[tracker.post_tool_use_hook]
  98. )
  99. ```
  100. The `parent_tool_use_id` links tool calls to their subagent:
  101. - Lead Agent spawns a Researcher via `Task` tool → gets ID "task_123"
  102. - All tool calls from that Researcher include `parent_tool_use_id = "task_123"`
  103. - Hooks use this ID to identify which subagent made the call
  104. ### Log Output
  105. **transcript.txt** - Human-readable:
  106. ```
  107. [RESEARCHER-1] → WebSearch
  108. Input: query='quantum computing 2025'
  109. [DATA-ANALYST-1] → Bash
  110. Input: python matplotlib chart generation
  111. ```
  112. **tool_calls.jsonl** - Structured JSON:
  113. ```json
  114. {"event":"tool_call_start","agent_id":"RESEARCHER-1","tool_name":"WebSearch",...}
  115. {"event":"tool_call_complete","success":true,"output_size":15234}
  116. ```

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.